Summary_Amyloids

Loading the necessary libraries and packages
Set paths to input and output
From raw measurements to estimated sample concentrations
Amyoids ELISA Results 2024-01-31
Amyoids ELISA Results 2024-02-08
Amyoids ELISA Results 2024-02-12

Loading the necessary libraries and packages

library(pacman)
pacman::p_load(data.table, ggplot2, lubridate, stringr, ggpubr, dplyr, cowplot, readxl, scales, knitr)

Set paths to input and output

Where do you store your all of the plate folders containing the excel files?
Output files will be stored in the same location!
Do you want to store the created figures here or somewhere else?

Input_Directory <- "./ALL_PLATES/"
save_to         <- Input_Directory

Add your cell line names for plots

In the lab we all know that cl204 is our triple KO cell line that lacks MyD88/IRAK4/IRAK1. However, it is likely cryptic to other readers. In addition, many of us use slightly different names, but mean the same thing.. To make plots that can be published and unify all sorts of synonyms, you may want to rename your cell lines in a reproducible manner. For this purpose, take a look at the ELISA_CL_KEY.csv and add your names!

CELL_LINE contains the name used on the plates
CL_NUMBER specifies the cell line number in our system
CL_NAME_ON_PLOT is equivalent to what you want to print in figures
PURPOSE is optional and just designates whether the cell line is used as a positive/negative/XYZ control or just a sample
INFO can be used to add additional information
PLOTTING_COLOR can be used to alter the color of your cell lines in the figures (not necessary, but it can add a nice touch if you wish)
ORDER_NO is used to plot your cell lines in the correct order
The most important cell lines against which we compare (the positive controls) are ranked highly (1 for WT, 2 for cl069, 3 for 3E10)
Negative controls follow (order number 4 and lower)
Samples come last - order them as you wish (I have continuously numbered them - synonyms can carry the same number)

# adjust the names as needed!
NAME_KEY <- fread("./ELISA_CL_KEY.csv", header = T) 
head(NAME_KEY, n = 10)

CELL_LINE	CL_NUMBER	CL_NAME_ON_PLOT	PURPOSE	INFO	PLOTTING_COLOR	ORDER_NO
EL4	cl011	WT_EL4	CONTROL	NA	#E00368	1
WT	cl011	WT_EL4	CONTROL	NA	#E00368	1
WT_EL4	cl011	WT_EL4	CONTROL	NA	#E00368	1
wt_Mph	cl051	WT_Mph	CONTROL	NA	#E00368	1
69	cl069	cl069	CONTROL	NA	salmon	2
cl069	cl069	cl069	CONTROL	NA	salmon	2
3E 10	cl028	MyD88-GFP	CONTROL	NA	salmon	3
3E10_GFP	cl028	MyD88-GFP	CONTROL	NA	salmon	3
MyD88-GFP	cl028	MyD88-GFP	CONTROL	NA	salmon	3
cl204	cl204	tKO_EL4	CONTROL	NA	#37aabd	4

We will need several functions to analyze the data. You do not need to change this.

source(file = ifelse(exists("https://raw.githubusercontent.com/tlobnow/coding_universe/main/scripts/ELISA_Fx.R"), 
                     yes =  "https://raw.githubusercontent.com/tlobnow/coding_universe/main/scripts/ELISA_Fx.R",
                     no  =  "./scripts/ELISA_Fx.R"))

From raw measurements to estimated sample concentrations

Calculate the mean values per Standard Curve dilution step
Plot the Standard curve and fit a linear trend line
Save the plots

We use the equation to estimate IL-2 conc. of our unknown samples

Run ELISA_Fx() to generate standard curves and calculate the IL-2 concentrations.

## [1] "Plates exist!"
## [1] "Secretion = slope*Intensity"
## [1] "Secretion = 352.419126169057*Intensity"

## [1] "Secretion = slope*Intensity"
## [1] "Secretion = 320.070381583682*Intensity"

## [1] "Secretion = slope*Intensity"
## [1] "Secretion = 306.346853473382*Intensity"

## [1] "Secretion = slope*Intensity"
## [1] "Secretion = 358.251952014523*Intensity"

## [1] "Secretion = slope*Intensity"
## [1] "Secretion = 298.804571083167*Intensity"

## [1] "Secretion = slope*Intensity"
## [1] "Secretion = 347.167928535214*Intensity"

## [1] "Secretion = slope*Intensity"
## [1] "Secretion = 559.423714931026*Intensity"

## [1] "Secretion = slope*Intensity"
## [1] "Secretion = 58.1684699766724*Intensity"

## [1] "Secretion = slope*Intensity"
## [1] "Secretion = 205.43426617438*Intensity"

## [1] "Secretion = slope*Intensity"
## [1] "Secretion = 1080.06019989381*Intensity"

## [1] "Secretion = slope*Intensity"
## [1] "Secretion = 1076.81100367551*Intensity"

## [1] "Secretion = slope*Intensity"
## [1] "Secretion = 659.569606790955*Intensity"

## [1] "Secretion = slope*Intensity"
## [1] "Secretion = 1015.10890429452*Intensity"

## [1] "Secretion = slope*Intensity"
## [1] "Secretion = 1076.81100367551*Intensity"

## [1] "Secretion = slope*Intensity"
## [1] "Secretion = 590.8221002662*Intensity"

Amyoids ELISA Results 2024-01-31

FILTER_VALUES = as.IDate("2024-01-31")
FILTER_VALUE = as.IDate("2024-01-31")
FILTER_TYPE   = "DATE"
POSITIVE_CTRL = c("WT_EL4", "cl069")
NEGATIVE_CTRL = c("204_TRIPLE_KO", "tKO_EL4")
save = T

x_label = ""
y_label = "relative IL-2 conc."
plot_title = "IL-2 ELISA"
subtitle = paste0("Assay performed for ", FILTER_VALUES, ", based on filter type ", FILTER_TYPE)
DF = All_plates_data
SEED = 1
plot_pval = T
plot_faceted_by_date = F
formatted_date <- format(as.IDate(FILTER_VALUES), "%Y%m%d")

# processing ELISA
ALL_COHORT_DATA <- lapply(FILTER_VALUES, function(FILTER_VALUE) {
  # Step 1: Filter & Subset the main data into lists
  COHORT_DATA <- filter_data(DF, FILTER_VALUE, FILTER_TYPE, POSITIVE_CTRL, NEGATIVE_CTRL)
  
  # extra filter
  COHORT_DATA <- COHORT_DATA %>%
  filter((Plate == 1 & STIM_CONCENTRATION == 5) | 
         (Plate == 2 & CELL_LINE == "cl204" & STIM_CONCENTRATION == 5) | 
         !(Plate == 2 & CELL_LINE == "cl204" & STIM_CONCENTRATION == 5 & MEASUREMENT > 1))
  
  COHORT_DATA <- COHORT_DATA %>%
    filter(STIM_CONCENTRATION == 5) %>%
    filter(!(CELL_LINE == "cl204" & MEASUREMENT > 1)) %>%
    filter(Plate != 3)
  
  # Prepare Normalization
  group_vars <- c("Date", "STIM_DAY")
  
  baseline <- COHORT_DATA %>%
      group_by(!!!syms(group_vars)) %>%
      filter((CELL_LINE %in% NEGATIVE_CTRL & CONDITION == "UNSTIM") | (CL_NAME_ON_PLOT %in% NEGATIVE_CTRL & CONDITION == "UNSTIM")) %>%
      summarise(baseline_control_value = mean(Concentration))
  
   # Join the calculated values with the dataset
  data <- left_join(COHORT_DATA, baseline, by = group_vars) %>%
      mutate(Concentration_REDUCED = case_when(!is.na(baseline_control_value) ~ Concentration - baseline_control_value, TRUE ~ Concentration))
  
  control_mean_per_day <- data %>%
      filter((CELL_LINE %in% POSITIVE_CTRL & CONDITION == "STIM") | (CL_NAME_ON_PLOT %in% POSITIVE_CTRL & CONDITION == "STIM")) %>%
      group_by(!!!syms(group_vars)) %>%
      summarise(control_mean_MEASUREMENT = case_when(mean(Concentration_REDUCED) > 0 ~ mean(Concentration_REDUCED), TRUE ~ -Inf))
    
    # Join the calculated control means
    data <- left_join(data, control_mean_per_day, by = group_vars)
    
    # Perform normalization
    DATA_NORMALIZED <- data %>%
      group_by(!!!syms(group_vars), CELL_LINE, CONDITION) %>%
      mutate(Concentration_NORMALIZED = case_when(Concentration_REDUCED / control_mean_MEASUREMENT < 0 ~ 0, TRUE ~ Concentration_REDUCED / control_mean_MEASUREMENT),
             triplicate_mean_per_day  = mean(Concentration_NORMALIZED)) %>%
      ungroup()
  NORMALIZED_DATA <- DATA_NORMALIZED
  return(NORMALIZED_DATA)
})

# Step 4: Combine the subset lists into one data frame
COMBINED_DATA <- bind_rows(ALL_COHORT_DATA)

# Step 5: Calculate the means per cohort
MEANS <- COMBINED_DATA %>%
  group_by(CELL_LINE, CONDITION, Date, PLOTTING_COLOR) %>%
  distinct(triplicate_mean_per_day, STIM_DAY, .keep_all = TRUE) %>%
  ungroup()

# Create a data frame with unique CL_NAME_ON_PLOT and corresponding PLOTTING_COLOR
color_mapping <- unique(MEANS[, c("CL_NAME_ON_PLOT", "PLOTTING_COLOR")])

# Filter rows where CONDITION is "UNSTIM" and set PLOTTING_COLOR to "gray50"
MEANS$PLOTTING_COLOR[MEANS$CONDITION == "UNSTIM"] <- "gray50"
# MEANS$PLOTTING_COLOR[MEANS$CONDITION == "STIM"] <- "#A6529A"

# Merge the MEANS data frame with the color_mapping data frame using dplyr
MEANS_merged <- MEANS %>%
  left_join(color_mapping) %>%
  mutate(CL_NAME_ON_PLOT = factor(CL_NAME_ON_PLOT, levels = unique(MEANS$CL_NAME_ON_PLOT)))

# Reorder CL_NAME_ON_PLOT based on ORDER_NO
MEANS_merged$CL_NAME_ON_PLOT <- reorder(MEANS$CL_NAME_ON_PLOT, -MEANS_merged$ORDER_NO)
MEANS_merged$STIM_DAY <- as.factor(MEANS_merged$STIM_DAY)

# Step 6: Calculate the mean of means (MOM) per cohort
MOM_SUBSET <- MEANS_merged %>%
  group_by(CELL_LINE, CL_NAME_ON_PLOT, CONDITION, Date, PLOTTING_COLOR, ORDER_NO) %>%
  summarise(triplicate_sd_per_day = sd(triplicate_mean_per_day),
            triplicate_mean_per_day = mean(triplicate_mean_per_day)) %>%
  distinct(.keep_all = TRUE)

# Step 7: Perform statistics per cohort
STATISTICAL_RESULTS <- perform_statistical_analysis(MEANS_merged, "CL_NAME_ON_PLOT")

# Your ggplot code
my_plot <- ggplot(MEANS_merged, aes(x = CL_NAME_ON_PLOT)) +
  geom_col(data = MOM_SUBSET,
           aes(y = triplicate_mean_per_day, fill = CONDITION), 
           position = position_dodge(width = 1),
           alpha = 0.5) +
  geom_point(data = MEANS_merged,
             aes(y = triplicate_mean_per_day, group = CONDITION, shape = STIM_DAY),
             position = position_jitterdodge(jitter.height = 0, jitter.width = 1.2, seed = SEED),
             col = "white", size = 4) +
  geom_point(data = MEANS_merged, aes(y = triplicate_mean_per_day, group = CONDITION, shape = STIM_DAY),
             position = position_jitterdodge(jitter.height = 0, jitter.width = 1.2, seed = SEED),
             col = "black", size = 3) +
  geom_errorbar(data = MOM_SUBSET, aes(ymin = triplicate_mean_per_day - triplicate_sd_per_day,
                                       ymax = triplicate_mean_per_day + triplicate_sd_per_day, group = CONDITION),
                width = 0.25, position = position_dodge(width = 1)) +
  labs(x = x_label,
       y = y_label) +
  scale_fill_manual(values = c("UNSTIM" = "gray50", "STIM" = "#A6529A")) +
  ggtitle(plot_title, paste0("STIM_CONCENTRATION = ", unique(MEANS_merged$STIM_CONCENTRATION), "ng/µL \n",
                                        "STIM_TIME = ", unique(MEANS_merged$STIM_TIME), "hrs")) +
  theme_cowplot() +
  theme(legend.position = "bottom") +
  theme(plot.title = element_text(hjust = 0.5)) +
  coord_flip() +
  facet_wrap(~Date)

my_plot

if (save) {
  # Step 9: Create Output Folder
  folder_path <- file.path(save_to, paste0("Plate_1_", formatted_date))
  if (!dir.exists(folder_path)) {dir.create(folder_path, recursive = TRUE)}   # Check if the folder exists, and if not, create it
  
  # Step 10: Save Plots and Data
  save_plots(folder_name = basename(folder_path), plots = list(my_plot))
  fwrite(COMBINED_DATA,       file = paste0(folder_path, "/COMBINED_DATA.csv"),       append = F, quote = F, row.names = F, col.names = T)
  fwrite(MEANS,               file = paste0(folder_path, "/MEANS.csv"),               append = F, quote = F, row.names = F, col.names = T)
  fwrite(MOM_SUBSET,          file = paste0(folder_path, "/MOM_SUBSET.csv"),          append = F, quote = F, row.names = F, col.names = T)
  fwrite(STATISTICAL_RESULTS, file = paste0(folder_path, "/STATISTICAL_RESULTS.csv"), append = F, quote = F, row.names = F, col.names = T)
}

FILTER_VALUES = as.IDate("2024-01-31")
FILTER_VALUE = as.IDate("2024-01-31")
FILTER_TYPE   = "DATE"
POSITIVE_CTRL = c("WT_EL4", "cl069")
NEGATIVE_CTRL = c("204_TRIPLE_KO", "tKO_EL4")
save = T

x_label = ""
y_label = "relative IL-2 conc."
plot_title = "IL-2 ELISA"
subtitle = paste0("Assay performed for ", FILTER_VALUES, ", based on filter type ", FILTER_TYPE)
DF = All_plates_data
SEED = 1
plot_pval = T
plot_faceted_by_date = F
formatted_date <- format(as.IDate(FILTER_VALUES), "%Y%m%d")

# processing ELISA
ALL_COHORT_DATA <- lapply(FILTER_VALUES, function(FILTER_VALUE) {
  # Step 1: Filter & Subset the main data into lists
  COHORT_DATA <- filter_data(DF, FILTER_VALUE, FILTER_TYPE, POSITIVE_CTRL, NEGATIVE_CTRL)
  
  # extra filter
  COHORT_DATA <- COHORT_DATA %>%
    filter(STIM_CONCENTRATION == 100) %>%
    filter(!(Plate == 2 & CELL_LINE == "cl204" & STIM_CONCENTRATION == 100 & MEASUREMENT > 1))

  # Prepare Normalization
  group_vars <- c("Date", "STIM_DAY")
  
  baseline <- COHORT_DATA %>%
      group_by(!!!syms(group_vars)) %>%
      filter((CELL_LINE %in% NEGATIVE_CTRL & CONDITION == "UNSTIM") | (CL_NAME_ON_PLOT %in% NEGATIVE_CTRL & CONDITION == "UNSTIM")) %>%
      summarise(baseline_control_value = mean(Concentration))
  
   # Join the calculated values with the dataset
  data <- left_join(COHORT_DATA, baseline, by = group_vars) %>%
      mutate(Concentration_REDUCED = case_when(!is.na(baseline_control_value) ~ Concentration - baseline_control_value, TRUE ~ Concentration))
  
  control_mean_per_day <- data %>%
      filter((CELL_LINE %in% POSITIVE_CTRL & CONDITION == "STIM") | (CL_NAME_ON_PLOT %in% POSITIVE_CTRL & CONDITION == "STIM")) %>%
      group_by(!!!syms(group_vars)) %>%
      summarise(control_mean_MEASUREMENT = case_when(mean(Concentration_REDUCED) > 0 ~ mean(Concentration_REDUCED), TRUE ~ -Inf))
    
    # Join the calculated control means
    data <- left_join(data, control_mean_per_day, by = group_vars)
    
    # Perform normalization
    DATA_NORMALIZED <- data %>%
      group_by(!!!syms(group_vars), CELL_LINE, CONDITION) %>%
      mutate(Concentration_NORMALIZED = case_when(Concentration_REDUCED / control_mean_MEASUREMENT < 0 ~ 0, TRUE ~ Concentration_REDUCED / control_mean_MEASUREMENT),
             triplicate_mean_per_day  = mean(Concentration_NORMALIZED)) %>%
      ungroup()
  NORMALIZED_DATA <- DATA_NORMALIZED
  return(NORMALIZED_DATA)
})

# Step 4: Combine the subset lists into one data frame
COMBINED_DATA <- bind_rows(ALL_COHORT_DATA)

# Step 5: Calculate the means per cohort
MEANS <- COMBINED_DATA %>%
  group_by(CELL_LINE, CONDITION, Date, PLOTTING_COLOR) %>%
  distinct(triplicate_mean_per_day, STIM_DAY, .keep_all = TRUE) %>%
  ungroup()

# Create a data frame with unique CL_NAME_ON_PLOT and corresponding PLOTTING_COLOR
color_mapping <- unique(MEANS[, c("CL_NAME_ON_PLOT", "PLOTTING_COLOR")])

# Filter rows where CONDITION is "UNSTIM" and set PLOTTING_COLOR to "gray50"
MEANS$PLOTTING_COLOR[MEANS$CONDITION == "UNSTIM"] <- "gray50"
# MEANS$PLOTTING_COLOR[MEANS$CONDITION == "STIM"] <- "#A6529A"

# Merge the MEANS data frame with the color_mapping data frame using dplyr
MEANS_merged <- MEANS %>%
  left_join(color_mapping) %>%
  mutate(CL_NAME_ON_PLOT = factor(CL_NAME_ON_PLOT, levels = unique(MEANS$CL_NAME_ON_PLOT)))

# Reorder CL_NAME_ON_PLOT based on ORDER_NO
MEANS_merged$CL_NAME_ON_PLOT <- reorder(MEANS$CL_NAME_ON_PLOT, -MEANS_merged$ORDER_NO)
MEANS_merged$STIM_DAY <- as.factor(MEANS_merged$STIM_DAY)

# Step 6: Calculate the mean of means (MOM) per cohort
MOM_SUBSET <- MEANS_merged %>%
  group_by(CELL_LINE, CL_NAME_ON_PLOT, CONDITION, Date, PLOTTING_COLOR, ORDER_NO) %>%
  summarise(triplicate_sd_per_day = sd(triplicate_mean_per_day),
            triplicate_mean_per_day = mean(triplicate_mean_per_day)) %>%
  distinct(.keep_all = TRUE)

# Step 7: Perform statistics per cohort
STATISTICAL_RESULTS <- perform_statistical_analysis(MEANS_merged, "CL_NAME_ON_PLOT")

# Your ggplot code
my_plot <- ggplot(MEANS_merged, aes(x = CL_NAME_ON_PLOT)) +
  geom_col(data = MOM_SUBSET,
           aes(y = triplicate_mean_per_day, fill = CONDITION), 
           position = position_dodge(width = 1),
           alpha = 0.5) +
  geom_point(data = MEANS_merged,
             aes(y = triplicate_mean_per_day, group = CONDITION, shape = STIM_DAY),
             position = position_jitterdodge(jitter.height = 0, jitter.width = 1.2, seed = SEED),
             col = "white", size = 4) +
  geom_point(data = MEANS_merged, aes(y = triplicate_mean_per_day, group = CONDITION, shape = STIM_DAY),
             position = position_jitterdodge(jitter.height = 0, jitter.width = 1.2, seed = SEED),
             col = "black", size = 3) +
  geom_errorbar(data = MOM_SUBSET, aes(ymin = triplicate_mean_per_day - triplicate_sd_per_day,
                                       ymax = triplicate_mean_per_day + triplicate_sd_per_day, group = CONDITION),
                width = 0.25, position = position_dodge(width = 1)) +
  labs(x = x_label,
       y = y_label) +
  scale_fill_manual(values = c("UNSTIM" = "gray50", "STIM" = "#A6529A")) +
  ggtitle(plot_title, paste0("STIM_CONCENTRATION = ", unique(MEANS_merged$STIM_CONCENTRATION), "ng/µL \n",
                                        "STIM_TIME = ", unique(MEANS_merged$STIM_TIME), "hrs")) +
  theme_cowplot() +
  theme(legend.position = "bottom") +
  theme(plot.title = element_text(hjust = 0.5)) +
  coord_flip() +
  facet_wrap(~Date)

my_plot

if (save) {
  # Step 9: Create Output Folder
  folder_path <- file.path(save_to, paste0("Plate_2_", formatted_date))
  if (!dir.exists(folder_path)) {dir.create(folder_path, recursive = TRUE)}   # Check if the folder exists, and if not, create it
  
  # Step 10: Save Plots and Data
  save_plots(folder_name = basename(folder_path), plots = list(my_plot))
  fwrite(COMBINED_DATA,       file = paste0(folder_path, "/COMBINED_DATA.csv"),       append = F, quote = F, row.names = F, col.names = T)
  fwrite(MEANS,               file = paste0(folder_path, "/MEANS.csv"),               append = F, quote = F, row.names = F, col.names = T)
  fwrite(MOM_SUBSET,          file = paste0(folder_path, "/MOM_SUBSET.csv"),          append = F, quote = F, row.names = F, col.names = T)
  fwrite(STATISTICAL_RESULTS, file = paste0(folder_path, "/STATISTICAL_RESULTS.csv"), append = F, quote = F, row.names = F, col.names = T)
}

FILTER_VALUES = as.IDate("2024-01-31")
FILTER_TYPE   = "DATE"
POSITIVE_CTRL = c("WT_EL4", "cl069")
NEGATIVE_CTRL = c("204_TRIPLE_KO", "tKO_EL4")
x_label = ""
y_label = "relative IL-2 conc."
plot_title = "IL-2 ELISA"
subtitle = paste0("Assay performed for ", FILTER_VALUES, ", based on filter type ", FILTER_TYPE)
save = T


SEED = 1
plot_pval = T
plot_faceted_by_date = F
formatted_date <- format(as.IDate(FILTER_VALUES), "%Y%m%d")

# processing ELISA
ALL_COHORT_DATA <- lapply(FILTER_VALUES, function(FILTER_VALUE) {
  # Step 1: Filter & Subset the main data into lists
  COHORT_DATA <- filter_data(DF, FILTER_VALUE, FILTER_TYPE, POSITIVE_CTRL, NEGATIVE_CTRL)
  
  # extra filter
  COHORT_DATA <- COHORT_DATA %>% filter(Plate == 3)
  
  # Step 2 and Step 3
  NORMALIZED_DATA <- calculate_baseline_and_control(COHORT_DATA, FILTER_TYPE, POSITIVE_CTRL, NEGATIVE_CTRL)
  return(NORMALIZED_DATA)
})

# Step 4: Combine the subset lists into one data frame
COMBINED_DATA <- bind_rows(ALL_COHORT_DATA)


# Step 5: Calculate the means per cohort
MEANS <- COMBINED_DATA %>%
  group_by(CELL_LINE, CONDITION, Date, PLOTTING_COLOR) %>%
  distinct(triplicate_mean_per_day, STIM_DAY, .keep_all = TRUE) %>%
  ungroup()

# Create a data frame with unique CL_NAME_ON_PLOT and corresponding PLOTTING_COLOR
color_mapping <- unique(MEANS[, c("CL_NAME_ON_PLOT", "PLOTTING_COLOR")])

# Filter rows where CONDITION is "UNSTIM" and set PLOTTING_COLOR to "gray50"
MEANS$PLOTTING_COLOR[MEANS$CONDITION == "UNSTIM"] <- "gray50"
# MEANS$PLOTTING_COLOR[MEANS$CONDITION == "STIM"] <- "#A6529A"

# Merge the MEANS data frame with the color_mapping data frame using dplyr
MEANS_merged <- MEANS %>%
  left_join(color_mapping) %>%
  mutate(CL_NAME_ON_PLOT = factor(CL_NAME_ON_PLOT, levels = unique(MEANS$CL_NAME_ON_PLOT)))

# Reorder CL_NAME_ON_PLOT based on ORDER_NO
MEANS_merged$CL_NAME_ON_PLOT <- reorder(MEANS$CL_NAME_ON_PLOT, -MEANS_merged$ORDER_NO)
MEANS_merged$STIM_DAY <- as.factor(MEANS_merged$STIM_DAY)

# Step 6: Calculate the mean of means (MOM) per cohort
MOM_SUBSET <- MEANS_merged %>%
  group_by(CELL_LINE, CL_NAME_ON_PLOT, CONDITION, Date, PLOTTING_COLOR, ORDER_NO) %>%
  summarise(triplicate_sd_per_day = sd(triplicate_mean_per_day),
            triplicate_mean_per_day = mean(triplicate_mean_per_day)) %>%
  distinct(.keep_all = TRUE)

# Step 7: Perform statistics per cohort
STATISTICAL_RESULTS <- perform_statistical_analysis(MEANS_merged, "CL_NAME_ON_PLOT")

# Your ggplot code
my_plot <- ggplot(MEANS_merged, aes(x = CL_NAME_ON_PLOT)) +
  geom_col(data = MOM_SUBSET,
           aes(y = triplicate_mean_per_day, fill = CONDITION), 
           position = position_dodge(width = 1),
           alpha = 0.5) +
  geom_point(data = MEANS_merged,
             aes(y = triplicate_mean_per_day, group = CONDITION, shape = STIM_DAY),
             position = position_jitterdodge(jitter.height = 0, jitter.width = 1.2, seed = SEED),
             col = "white", size = 4) +
  geom_point(data = MEANS_merged, aes(y = triplicate_mean_per_day, group = CONDITION, shape = STIM_DAY),
             position = position_jitterdodge(jitter.height = 0, jitter.width = 1.2, seed = SEED),
             col = "black", size = 3) +
  geom_errorbar(data = MOM_SUBSET, aes(ymin = triplicate_mean_per_day - triplicate_sd_per_day,
                                       ymax = triplicate_mean_per_day + triplicate_sd_per_day, group = CONDITION),
                width = 0.25, position = position_dodge(width = 1)) +
  labs(x = x_label,
       y = y_label) +
  scale_fill_manual(values = c("UNSTIM" = "gray50", "STIM" = "#A6529A")) +
  ggtitle(plot_title, paste0("STIM_CONCENTRATION = ", unique(MEANS_merged$STIM_CONCENTRATION), "ng/µL \n",
                                        "STIM_TIME = ", unique(MEANS_merged$STIM_TIME), "hrs")) +
  theme_cowplot() +
  theme(legend.position = "bottom") +
  theme(plot.title = element_text(hjust = 0.5)) +
  coord_flip() +
  facet_wrap(~Date)

my_plot

if (save) {
  # Step 9: Create Output Folder
  folder_path <- file.path(save_to, paste0("Plate_3_", formatted_date))
  if (!dir.exists(folder_path)) {dir.create(folder_path, recursive = TRUE)}   # Check if the folder exists, and if not, create it
  
  # Step 10: Save Plots and Data
  save_plots(folder_name = basename(folder_path), plots = list(my_plot))
  fwrite(COMBINED_DATA,       file = paste0(folder_path, "/COMBINED_DATA.csv"),       append = F, quote = F, row.names = F, col.names = T)
  fwrite(MEANS,               file = paste0(folder_path, "/MEANS.csv"),               append = F, quote = F, row.names = F, col.names = T)
  fwrite(MOM_SUBSET,          file = paste0(folder_path, "/MOM_SUBSET.csv"),          append = F, quote = F, row.names = F, col.names = T)
  fwrite(STATISTICAL_RESULTS, file = paste0(folder_path, "/STATISTICAL_RESULTS.csv"), append = F, quote = F, row.names = F, col.names = T)
}

Amyoids ELISA Results 2024-02-08

FILTER_VALUES = as.IDate("2024-02-08")
FILTER_TYPE   = "DATE"
POSITIVE_CTRL = c("WT_EL4", "cl069")
NEGATIVE_CTRL = c("204_TRIPLE_KO", "tKO_EL4")
x_label = ""
y_label = "relative IL-2 conc."
plot_title = "IL-2 ELISA"
subtitle = paste0("Assay performed for ", FILTER_VALUES, ", based on filter type ", FILTER_TYPE)
save = T


SEED = 1
plot_pval = T
plot_faceted_by_date = F
formatted_date <- format(as.IDate(FILTER_VALUES), "%Y%m%d")

# processing ELISA
ALL_COHORT_DATA <- lapply(FILTER_VALUES, function(FILTER_VALUE) {
  # Step 1: Filter & Subset the main data into lists
  COHORT_DATA <- filter_data(DF, FILTER_VALUE, FILTER_TYPE, POSITIVE_CTRL, NEGATIVE_CTRL)
  
  # extra filter
  COHORT_DATA <- COHORT_DATA %>% filter(Plate == 1)
  
  # Step 2 and Step 3
  NORMALIZED_DATA <- calculate_baseline_and_control(COHORT_DATA, FILTER_TYPE, POSITIVE_CTRL, NEGATIVE_CTRL)
  return(NORMALIZED_DATA)
})

# Step 4: Combine the subset lists into one data frame
COMBINED_DATA <- bind_rows(ALL_COHORT_DATA)

# Step 5: Calculate the means per cohort
MEANS <- COMBINED_DATA %>%
  group_by(CELL_LINE, CONDITION, Date, PLOTTING_COLOR) %>%
  distinct(triplicate_mean_per_day, STIM_DAY, .keep_all = TRUE) %>%
  ungroup()

# Create a data frame with unique CL_NAME_ON_PLOT and corresponding PLOTTING_COLOR
color_mapping <- unique(MEANS[, c("CL_NAME_ON_PLOT", "PLOTTING_COLOR")])

# Filter rows where CONDITION is "UNSTIM" and set PLOTTING_COLOR to "gray50"
MEANS$PLOTTING_COLOR[MEANS$CONDITION == "UNSTIM"] <- "gray50"
# MEANS$PLOTTING_COLOR[MEANS$CONDITION == "STIM"] <- "#A6529A"

# Merge the MEANS data frame with the color_mapping data frame using dplyr
MEANS_merged <- MEANS %>%
  left_join(color_mapping) %>%
  mutate(CL_NAME_ON_PLOT = factor(CL_NAME_ON_PLOT, levels = unique(MEANS$CL_NAME_ON_PLOT)))

# Reorder CL_NAME_ON_PLOT based on ORDER_NO
MEANS_merged$CL_NAME_ON_PLOT <- reorder(MEANS$CL_NAME_ON_PLOT, -MEANS_merged$ORDER_NO)
MEANS_merged$STIM_DAY <- as.factor(MEANS_merged$STIM_DAY)

# Step 6: Calculate the mean of means (MOM) per cohort
MOM_SUBSET <- MEANS_merged %>%
  group_by(CELL_LINE, CL_NAME_ON_PLOT, CONDITION, Date, PLOTTING_COLOR, ORDER_NO) %>%
  summarise(triplicate_sd_per_day = sd(triplicate_mean_per_day),
            triplicate_mean_per_day = mean(triplicate_mean_per_day)) %>%
  distinct(.keep_all = TRUE)

# Step 7: Perform statistics per cohort
STATISTICAL_RESULTS <- perform_statistical_analysis(MEANS_merged, "CL_NAME_ON_PLOT")

# Your ggplot code
my_plot <- ggplot(MEANS_merged, aes(x = CL_NAME_ON_PLOT)) +
  geom_col(data = MOM_SUBSET,
           aes(y = triplicate_mean_per_day, fill = CONDITION), 
           position = position_dodge(width = 1),
           alpha = 0.5) +
  geom_point(data = MEANS_merged,
             aes(y = triplicate_mean_per_day, group = CONDITION, shape = STIM_DAY),
             position = position_jitterdodge(jitter.height = 0, jitter.width = 1.2, seed = SEED),
             col = "white", size = 4) +
  geom_point(data = MEANS_merged, aes(y = triplicate_mean_per_day, group = CONDITION, shape = STIM_DAY),
             position = position_jitterdodge(jitter.height = 0, jitter.width = 1.2, seed = SEED),
             col = "black", size = 3) +
  geom_errorbar(data = MOM_SUBSET, aes(ymin = triplicate_mean_per_day - triplicate_sd_per_day,
                                       ymax = triplicate_mean_per_day + triplicate_sd_per_day, group = CONDITION),
                width = 0.25, position = position_dodge(width = 1)) +
  labs(x = x_label,
       y = y_label) +
  scale_fill_manual(values = c("UNSTIM" = "gray50", "STIM" = "#A6529A")) +
  ggtitle(plot_title, subtitle = paste0("STIM_CONCENTRATION = ", unique(MEANS_merged$STIM_CONCENTRATION), "ng/µL \n",
                                        "STIM_TIME = ", unique(MEANS_merged$STIM_TIME), "hrs")) +
  theme_cowplot() +
  theme(legend.position = "bottom") +
  theme(plot.title = element_text(hjust = 0.5)) +
  coord_flip() +
  facet_wrap(~Date)

my_plot

if (save) {
  # Step 9: Create Output Folder
  folder_path <- file.path(save_to, paste0("Plate_1_", formatted_date))
  if (!dir.exists(folder_path)) {dir.create(folder_path, recursive = TRUE)}   # Check if the folder exists, and if not, create it
  
  # Step 10: Save Plots and Data
  save_plots(folder_name = basename(folder_path), plots = list(my_plot))
  fwrite(COMBINED_DATA,       file = paste0(folder_path, "/COMBINED_DATA.csv"),       append = F, quote = F, row.names = F, col.names = T)
  fwrite(MEANS,               file = paste0(folder_path, "/MEANS.csv"),               append = F, quote = F, row.names = F, col.names = T)
  fwrite(MOM_SUBSET,          file = paste0(folder_path, "/MOM_SUBSET.csv"),          append = F, quote = F, row.names = F, col.names = T)
  fwrite(STATISTICAL_RESULTS, file = paste0(folder_path, "/STATISTICAL_RESULTS.csv"), append = F, quote = F, row.names = F, col.names = T)
}

FILTER_VALUES = as.IDate("2024-02-08")
FILTER_TYPE   = "DATE"
POSITIVE_CTRL = c("WT_EL4", "cl069")
NEGATIVE_CTRL = c("204_TRIPLE_KO", "tKO_EL4")
x_label = ""
y_label = "relative IL-2 conc."
plot_title = "IL-2 ELISA"
subtitle = paste0("Assay performed for ", FILTER_VALUES, ", based on filter type ", FILTER_TYPE)
save = T


SEED = 600
plot_pval = T
plot_faceted_by_date = F
formatted_date <- format(as.IDate(FILTER_VALUES), "%Y%m%d")

# processing ELISA
ALL_COHORT_DATA <- lapply(FILTER_VALUES, function(FILTER_VALUE) {
  # Step 1: Filter & Subset the main data into lists
  COHORT_DATA <- filter_data(DF, FILTER_VALUE, FILTER_TYPE, POSITIVE_CTRL, NEGATIVE_CTRL)
  
  # extra filter
  COHORT_DATA <- COHORT_DATA %>% filter(Plate == 2)
  
  # Step 2 and Step 3
  NORMALIZED_DATA <- calculate_baseline_and_control(COHORT_DATA, FILTER_TYPE, POSITIVE_CTRL, NEGATIVE_CTRL)
  return(NORMALIZED_DATA)
})

# Step 4: Combine the subset lists into one data frame
COMBINED_DATA <- bind_rows(ALL_COHORT_DATA)

# Step 5: Calculate the means per cohort
MEANS <- COMBINED_DATA %>%
  group_by(CELL_LINE, CONDITION, Date, PLOTTING_COLOR) %>%
  distinct(triplicate_mean_per_day, STIM_DAY, .keep_all = TRUE) %>%
  ungroup()

# Create a data frame with unique CL_NAME_ON_PLOT and corresponding PLOTTING_COLOR
color_mapping <- unique(MEANS[, c("CL_NAME_ON_PLOT", "PLOTTING_COLOR")])

# Filter rows where CONDITION is "UNSTIM" and set PLOTTING_COLOR to "gray50"
MEANS$PLOTTING_COLOR[MEANS$CONDITION == "UNSTIM"] <- "gray50"
# MEANS$PLOTTING_COLOR[MEANS$CONDITION == "STIM"] <- "#A6529A"

# Merge the MEANS data frame with the color_mapping data frame using dplyr
MEANS_merged <- MEANS %>%
  left_join(color_mapping) %>%
  mutate(CL_NAME_ON_PLOT = factor(CL_NAME_ON_PLOT, levels = unique(MEANS$CL_NAME_ON_PLOT)))

# Reorder CL_NAME_ON_PLOT based on ORDER_NO
MEANS_merged$CL_NAME_ON_PLOT <- reorder(MEANS$CL_NAME_ON_PLOT, -MEANS_merged$ORDER_NO)
MEANS_merged$STIM_DAY <- as.factor(MEANS_merged$STIM_DAY)

# Step 6: Calculate the mean of means (MOM) per cohort
MOM_SUBSET <- MEANS_merged %>%
  group_by(CELL_LINE, CL_NAME_ON_PLOT, CONDITION, Date, PLOTTING_COLOR, ORDER_NO) %>%
  summarise(triplicate_sd_per_day = sd(triplicate_mean_per_day),
            triplicate_mean_per_day = mean(triplicate_mean_per_day)) %>%
  distinct(.keep_all = TRUE)

# Step 7: Perform statistics per cohort
STATISTICAL_RESULTS <- perform_statistical_analysis(MEANS_merged, "CL_NAME_ON_PLOT")

# Your ggplot code
my_plot <- ggplot(MEANS_merged, aes(x = CL_NAME_ON_PLOT)) +
  geom_col(data = MOM_SUBSET,
           aes(y = triplicate_mean_per_day, fill = CONDITION), 
           position = position_dodge(width = 1),
           alpha = 0.5) +
  geom_point(data = MEANS_merged,
             aes(y = triplicate_mean_per_day, group = CONDITION, shape = STIM_DAY),
             position = position_jitterdodge(jitter.height = 0, jitter.width = 1.2, seed = SEED),
             col = "white", size = 4) +
  geom_point(data = MEANS_merged, aes(y = triplicate_mean_per_day, group = CONDITION, shape = STIM_DAY),
             position = position_jitterdodge(jitter.height = 0, jitter.width = 1.2, seed = SEED),
             col = "black", size = 3) +
  geom_errorbar(data = MOM_SUBSET, aes(ymin = triplicate_mean_per_day - triplicate_sd_per_day,
                                       ymax = triplicate_mean_per_day + triplicate_sd_per_day, group = CONDITION),
                width = 0.25, position = position_dodge(width = 1)) +
  labs(x = x_label,
       y = y_label) +
  scale_fill_manual(values = c("UNSTIM" = "gray50", "STIM" = "#A6529A")) +
  ggtitle(plot_title, subtitle = paste0("STIM_CONCENTRATION = ", unique(MEANS_merged$STIM_CONCENTRATION), "ng/µL \n",
                                        "STIM_TIME = ", unique(MEANS_merged$STIM_TIME), "hrs")) +
  theme_cowplot() +
  theme(legend.position = "bottom") +
  theme(plot.title = element_text(hjust = 0.5)) +
  coord_flip() +
  facet_wrap(~Date)

my_plot

if (save) {
  # Step 9: Create Output Folder
  folder_path <- file.path(save_to, paste0("Plate_2_", formatted_date))
  if (!dir.exists(folder_path)) {dir.create(folder_path, recursive = TRUE)}   # Check if the folder exists, and if not, create it
  
  # Step 10: Save Plots and Data
  save_plots(folder_name = basename(folder_path), plots = list(my_plot))
  fwrite(COMBINED_DATA,       file = paste0(folder_path, "/COMBINED_DATA.csv"),       append = F, quote = F, row.names = F, col.names = T)
  fwrite(MEANS,               file = paste0(folder_path, "/MEANS.csv"),               append = F, quote = F, row.names = F, col.names = T)
  fwrite(MOM_SUBSET,          file = paste0(folder_path, "/MOM_SUBSET.csv"),          append = F, quote = F, row.names = F, col.names = T)
  fwrite(STATISTICAL_RESULTS, file = paste0(folder_path, "/STATISTICAL_RESULTS.csv"), append = F, quote = F, row.names = F, col.names = T)
}

Amyoids ELISA Results 2024-02-12

FILTER_VALUES = as.IDate("2024-02-12")
FILTER_TYPE   = "DATE"
POSITIVE_CTRL = c("WT_EL4", "cl069")
NEGATIVE_CTRL = c("204_TRIPLE_KO", "tKO_EL4")
x_label = ""
y_label = "relative IL-2 conc."
plot_title = "IL-2 ELISA"
subtitle = paste0("Assay performed for ", FILTER_VALUES, ", based on filter type ", FILTER_TYPE)
save = T


SEED = 600
plot_pval = T
plot_faceted_by_date = F
formatted_date <- format(as.IDate(FILTER_VALUES), "%Y%m%d")

# processing ELISA
ALL_COHORT_DATA <- lapply(FILTER_VALUES, function(FILTER_VALUE) {
  # Step 1: Filter & Subset the main data into lists
  COHORT_DATA <- filter_data(DF, FILTER_VALUE, FILTER_TYPE, POSITIVE_CTRL, NEGATIVE_CTRL)
  
  # Step 2 and Step 3
  NORMALIZED_DATA <- calculate_baseline_and_control(COHORT_DATA, FILTER_TYPE, POSITIVE_CTRL, NEGATIVE_CTRL)
  return(NORMALIZED_DATA)
})

## `summarise()` has grouped output by 'Date', 'STIM_DAY'. You can override using
## the `.groups` argument.
## `summarise()` has grouped output by 'Date', 'STIM_DAY'. You can override using
## the `.groups` argument.

# Step 4: Combine the subset lists into one data frame
COMBINED_DATA <- bind_rows(ALL_COHORT_DATA)

# Step 5: Calculate the means per cohort
MEANS <- COMBINED_DATA %>%
  group_by(CELL_LINE, CONDITION, Date, PLOTTING_COLOR) %>%
  distinct(triplicate_mean_per_day, STIM_DAY, .keep_all = TRUE) %>%
  ungroup()

# Create a data frame with unique CL_NAME_ON_PLOT and corresponding PLOTTING_COLOR
color_mapping <- unique(MEANS[, c("CL_NAME_ON_PLOT", "PLOTTING_COLOR")])

# Filter rows where CONDITION is "UNSTIM" and set PLOTTING_COLOR to "gray50"
MEANS$PLOTTING_COLOR[MEANS$CONDITION == "UNSTIM"] <- "gray50"
# MEANS$PLOTTING_COLOR[MEANS$CONDITION == "STIM"] <- "#A6529A"

# Merge the MEANS data frame with the color_mapping data frame using dplyr
MEANS_merged <- MEANS %>%
  left_join(color_mapping) %>%
  mutate(CL_NAME_ON_PLOT = factor(CL_NAME_ON_PLOT, levels = unique(MEANS$CL_NAME_ON_PLOT)))

## Joining with `by = join_by(CL_NAME_ON_PLOT, PLOTTING_COLOR)`

# Reorder CL_NAME_ON_PLOT based on ORDER_NO
MEANS_merged$CL_NAME_ON_PLOT <- reorder(MEANS$CL_NAME_ON_PLOT, -MEANS_merged$ORDER_NO)
MEANS_merged$STIM_DAY <- as.factor(MEANS_merged$STIM_DAY)

# Step 6: Calculate the mean of means (MOM) per cohort
MOM_SUBSET <- MEANS_merged %>%
  group_by(CELL_LINE, CL_NAME_ON_PLOT, CONDITION, Date, PLOTTING_COLOR, ORDER_NO) %>%
  summarise(triplicate_sd_per_day = sd(triplicate_mean_per_day),
            triplicate_mean_per_day = mean(triplicate_mean_per_day)) %>%
  distinct(.keep_all = TRUE)

## `summarise()` has grouped output by 'CELL_LINE', 'CL_NAME_ON_PLOT',
## 'CONDITION', 'Date', 'PLOTTING_COLOR'. You can override using the `.groups`
## argument.

# Step 7: Perform statistics per cohort
STATISTICAL_RESULTS <- perform_statistical_analysis(MEANS_merged, "CL_NAME_ON_PLOT")

# Your ggplot code
my_plot <- ggplot(MEANS_merged, aes(x = CL_NAME_ON_PLOT)) +
  geom_col(data = MOM_SUBSET,
           aes(y = triplicate_mean_per_day, fill = CONDITION), 
           position = position_dodge(width = 1),
           alpha = 0.5) +
  geom_point(data = MEANS_merged,
             aes(y = triplicate_mean_per_day, group = CONDITION, shape = STIM_DAY),
             position = position_jitterdodge(jitter.height = 0, jitter.width = 1.2, seed = SEED),
             col = "white", size = 4) +
  geom_point(data = MEANS_merged, aes(y = triplicate_mean_per_day, group = CONDITION, shape = STIM_DAY),
             position = position_jitterdodge(jitter.height = 0, jitter.width = 1.2, seed = SEED),
             col = "black", size = 3) +
  geom_errorbar(data = MOM_SUBSET, aes(ymin = triplicate_mean_per_day - triplicate_sd_per_day,
                                       ymax = triplicate_mean_per_day + triplicate_sd_per_day, group = CONDITION),
                width = 0.25, position = position_dodge(width = 1)) +
  labs(x = x_label,
       y = y_label) +
  scale_fill_manual(values = c("UNSTIM" = "gray50", "STIM" = "#A6529A")) +
  ggtitle(plot_title, subtitle = paste0("STIM_CONCENTRATION = ", unique(MEANS_merged$STIM_CONCENTRATION), "ng/µL \n",
                                        "STIM_TIME = ", unique(MEANS_merged$STIM_TIME), "hrs")) +
  theme_cowplot() +
  theme(legend.position = "bottom") +
  theme(plot.title = element_text(hjust = 0.5)) +
  coord_flip() +
  facet_wrap(~Date)

my_plot

## Warning: Removed 14 rows containing missing values (`geom_col()`).

## Warning: Removed 14 rows containing missing values (`geom_point()`).
## Removed 14 rows containing missing values (`geom_point()`).

if (save) {
  # Step 9: Create Output Folder
  folder_path <- file.path(save_to, paste0("Plate_1_", formatted_date))
  if (!dir.exists(folder_path)) {dir.create(folder_path, recursive = TRUE)}   # Check if the folder exists, and if not, create it
  
  # Step 10: Save Plots and Data
  save_plots(folder_name = basename(folder_path), plots = list(my_plot))
  fwrite(COMBINED_DATA,       file = paste0(folder_path, "/COMBINED_DATA.csv"),       append = F, quote = F, row.names = F, col.names = T)
  fwrite(MEANS,               file = paste0(folder_path, "/MEANS.csv"),               append = F, quote = F, row.names = F, col.names = T)
  fwrite(MOM_SUBSET,          file = paste0(folder_path, "/MOM_SUBSET.csv"),          append = F, quote = F, row.names = F, col.names = T)
  fwrite(STATISTICAL_RESULTS, file = paste0(folder_path, "/STATISTICAL_RESULTS.csv"), append = F, quote = F, row.names = F, col.names = T)
}

## Warning: Removed 14 rows containing missing values (`geom_col()`).
## Removed 14 rows containing missing values (`geom_point()`).
## Removed 14 rows containing missing values (`geom_point()`).

## Warning: Removed 14 rows containing missing values (`geom_col()`).

## Warning: Removed 14 rows containing missing values (`geom_point()`).
## Removed 14 rows containing missing values (`geom_point()`).